Thanks to Caleb Sykes awesome database. I created rankings for the PCH district. These are world rankings so they are vs. all 3653 teams with data.
The first column is a combined score based on win%, rocket%, and hab3% … the win% is counted twice. I wanted a way to give an aggregate ranking point score and this made sense to me.
The next column is that score as a percentile rank across all teams. Higher is better.
The other 3 columns are the constituent elements of this: win, rocket, and hab.
The last column is for total scored game pieces. In eliminations, completing a rocket doesn’t matter in scoring.
1746 did have the highest average for scored game pieces, but due to lesser performance in previous events, they didn’t make the top 5.
Based on just week 6 …
1102 was #5 globally in scored game pieces
1746 was #8 globally in scored game pieces
2974 was #12 in rocket RP
4910 was #7 in rocket RP
that [win] number confuses me too … we have a winning record for all our matches yet the number shown is around 0.49 … I used it anyway, which is something I wouldn’t do professionally.
So, the “calculated contribution to win” metric is an OPR-like calculation on the binary outcome of a win (1) or loss (0) for each match. It can thus be interpreted kind of like how much of a “percentage” of a win your team contributes. So since there’s one win per match to be distributed amongst 6 teams, the average value for this metric will be 1/6 = 0.17. It can roughly be converted to win rate by taking your value + 2*.17 (assume you have two average partners and average opponents). So 4910’s value of 0.40 corresponds approximately to a win rate of 0.74.
Thank you. So looking at TBA, 4910 is 44-12 (0.79) and 29-8 (0.78) if you just count qualification matches … so pretty accurate for these purposes
While none of the PCH teams made the top 25 by either method, I think PCH is in pretty good shape this year with 4 teams able to solo a rocket and some reliable L3 climbers who can play tough defense.
To go into a little more depth and more data . After re watching all of 1102’s, 1746’s, 2974’s, and 4910’s matches while taking data on hatches/cargo scored here is what I got.
Averages for all matches:
2974: 11.056 (Max: 16, Min: 4)
1746: 10.111 (Max: 14, Min: 8)
4910: 8.44 (Max: 12, Min: 4)
1102: 7.833 (Max: 12, Min 0)
When calculating scored game pieces I only counted what was scored and stayed scored. This means if a team scored a hatch and then a cargo nocked it out neither of them would be counted. If a team scored a cargo on another team’s hatch and it fell off the cargo would be counted. Scoring a second cargo ball was not counted. Placing a hatch on a cargo ball during sandstorm would be counted as a hatch and cargo (this became important during elems).
Impressive both the depth and tenacity! While 2974 came out on top, here is my qualitative assessment:
1746 and 2974 are non-L3 climbers specializing in the rocket. 1746 is the fastest hatch robot, due to automation, elevator speed, and driver. The cargo intake is just not as fast. 2974 is the opposite. While the hatch mechanism is pretty tolerant, we’re just not as fast at landing as we’d like (better vision will help). But the cargo intake is solid - we took some BIG hits without losing the ball. Wish we could climb!
1102 and 4910 are do-it-all robots. I think they are very similar in scoring speed and ability. 1102 is very fast with possibly more decisive driving. But I think 4910 may have more runway to score big. The weakness, and I say this with love, is the the long “boom” of 1102 which can make it take longer to land pieces. 4910 has the edge on climbing due to a more stable footing (when they don’t lose comms).
If we had the resources to build a second robot we would have shown up to state with an entirely different system to mount in place of our arm… now that we have scrapped the flip. The design team actually planned for something that looks a lot like what 1746 does.
I think we have a few small things we can tweak more going into worlds to make placement more consistent and fast. I know as the event went on this past weekend we were getting better as time went on at understanding how to play around it. We were able to finish a rocket with 30+ seconds left at least once. A lot of it has to do with our Limelight placement still isn’t ideal. We’re probably remounting it again, first thing once we get to worlds.
. After re watching all of 1102’s, 1746’s, 2974’s, and 4910’s matches while taking data on hatches/cargo scored here is what I got.
